Log in

View Full Version here: : TPoint development suggestion


montewilson
06-05-2007, 11:42 AM
Guys - Here is a post I put on the Software Bisque forum. I guess there may be a few people here using TPoint too, which is a godsend for goto users. If so I would be grateful for their input to this product development suggestion.


Like most people in this section I'm a happy TPoint user. It has made a wonderful difference to my pointing accuracy. After six or so points the star is always very close to the centre of the frame and only gets closer with additional points.

That said I think the product could be improved with respect to data entry, star centering ect.

Let me explain how I do it to give you an idea of where I am coming from.
I have an EM-200 with a Tak FSQ-106 and STL-11000m with which I use TheSky 6 & CCDSoft.
I use an STV as my "crosshairs" in E Finder mode so I don't have to look through an illuminated eyepiece. It works well. But one of my biggest frustrations is using the motion controls in TheSky. They can't be improved and work fine but centering the star with these controls is quite a challenge.
I was doing this last night and it occurred to me there is probably (from my layman's persepective) a way to use the existing faetures of CCDSoft and TheSky to automate the mapping process.

CCDSoft can calculate the position of and centre stars and take images on command. TheSky can talk to TPoint so I am guessing it should be possible to automate the mapping process.

My thought is the mapping command would exist in TheSky. Prior to telling TheSky to map CCD Soft would have to be running. CCSSoft would have to be calibrated just as for autoguiding. This will allow it to centre the star. The STL can be set to image at 9x9 to reduce dowload times. Other CDD's have similar features I'm sure.

Lets assume we start in a given constellation. After we sync on a bright star in the constellation and then tell to it start auto mapping, it will go for another star of a set minimum brightness within a set distance. We are assuming the mount is more or less polar aligned so if CCDSoft chooses a close star it should appear on the camera ccd . At this point the program assumes the brightest star in the field is the desired one and centres it and maps it. Once this is done The Sky chooses another nearby star and repeats the process. As the number of points mapped increases The Sky can choose more distant stars for subsequent mapping points.

If it chooses the wrong star in a crowded field and an error is detected by TPoint it can be ignored and the system chooses a new star and trys again.
This would be really helpful for people who are setting up in new places frequently. It can done when the sky is not quite ready for imaging so when it is there is not a lot of time spent mapping.

There is one thing I should mention. The motion controls in CCDSoft work (with the Tak) as a ghost of the hand controller so if it is set in guiding mode the correcting is very slow and centering a star could be time consuming. If the mount is in slew mode the star could rocket right past the centre point and become very difficult to get exactly right. TheSky can control movement very well at both speeds but to my knowledge doesn't control actions in CCDSoft.

Can TheSky get x-y values from CCDSoft and make corrections from there to take advantage of Sky's ability to control movement? This would be faster and more accurate.

In a perfect world this software could estimate the offset of the star on the ccd sensor and not need to centre it at all. This would speed things up even more.

I appreciate this all sounds easy to someone who doesn't program computers but if it isn't particularly difficult it would be a great addition and a strong selling point too.

jase
06-05-2007, 01:50 PM
I researched TheSky:TPoint:CCDSoft combination and didn't feel it was very well developed. Don't get me wrong, all three products are excellent, but the interoperability was of some concern even though they are developed by the one developer (Bisque).

I ended up going the opposite direction - MaximDL/CCD (which includes Pinpoint LE) and MaxPoint. As you're probably aware, MaxPoint is the equivalent of TPoint. What I liked about this combination was the ability for MaximDL to plate solve using PinPoint LE Astrometric Engine. Hence MaxPoint slewed to a star, MaximDL would take an image, plate solve it and then center it automatically. This would then be reflected in the MaxPoint pointing model. I've built a 120 star pointing model using this method with no problems to date.

The above is great, but what I really like is that every component is ASCOM compliant. MaxPoint acts as an ASCOM server (hub), so I simply point ASCOM clients such as - MaximDL/CCD,FocusMax,PolarAlignMax and TheSky6 (using the TeleAPI) to the hub. That way all programs benefit from the MaxPoint pointing model.

I recently downloaded the 6GB USNO-A2.0 catalog that contains 526,280,881 stars. I haven't had the time to build a larger pointing model with this catalog as yet, but it’s on the "to-do" list. I was actually impressed with the Losmandy Gemini Level 4 pointing model, but having used MaxPoint I realised the Gemini model doesn't deliver arc-sec accuracy across the sky. Not everyone requires this level of accuracy, but my goal is to establish robotic operations where I can image all night remotely. The observatory is two hours away from where I live. With no one at the scope to check pointing, software is my only option.

I think you're going to need a way to plate solve with CCDsoft to automate the process. You may wish to have a look at Pinpoint - http://pinpoint.dc3.com/

I'm assuming you're using the automapper, if not... you'll need this...
http://www.newastro.com/newastro/downloads/automap2/default.htm

rogerg
06-05-2007, 04:04 PM
Either of you guy's (or anyone else) have experience using TPoint with an LX200? I'd like to hear from people who have.

I have no problem doing plate solve's automatically in TheSky/CCDSoft, but that's using CCDCommander. My only problem is pointing accuracy, getting near the target to start with.

Roger.

montewilson
06-05-2007, 04:38 PM
Thanks Jase for the detailed reply. I will watch with interest to see what Bisque's response is and have a look at pinpoint.

As for the LX200 - no experience sorry.

jase
06-05-2007, 07:18 PM
Roger, I can't speak specifically for the LX200, but I've used the combination I mentioned above (Maxpoint:MaximDL/CCD/Pinpoint) with the Nexstar 11. I significantly improved the pointing, but it I found it still some what erroneous.
The issue you have with SCT's is the primary mirror. No matter who the manufacturer is or the age of the scope, the primary always experiences slight (in some cases large) mirror shifts as the scope slews across the meridian. Point modelling software can correct this, but only if the mirror shifts the precise amount every time (which is unlikely). Regardless, you will certainly improve the pointing accuracy of the LX200 using such software, though I can't commit on how much more accuracy you will obtain. Perhaps download a trial of TPoint or MaxPoint to see where you get.

rogerg
07-05-2007, 10:27 AM
Hmm, interesting.. thanks.

Vince G
09-05-2007, 11:35 AM
Roger,
If you are having pointing problems by not getting near the object to start with, then, using CCDCommander, try two plate solves. It should get you spot on.
Vince G

rogerg
09-05-2007, 01:44 PM
I use CCD Commander, but it doesn't help when the object is outside the camera's FOV all together :)

My camera's FOV is very small (13 arc min x 7 arc min - or there abouts). At last try, slewing short distances (up to 45 degree) isn't a problem, however generally the pointing accuracy isn't good enough to rely upon automated plate solve's, which is exactly my problem.

When my scope is up and running fully again I plan to correct the PA which will get it back to within the FOV all over the sky, but that still often isn't enough for a plate solve for my setup, because of the small FOV (at least that's all I can put it down to).

Roger.

montewilson
10-05-2007, 08:17 AM
OK Guys - here is where I admit that I should have looked a bit harder before I started this thread. Bisque actually have what I suggested (Damn I'm smart :D ) the only difference is that they have had it for some years and its is far better and more sophisticated than that which I conjected.

They use plate solve and it will do automatic mapping across the whole sky based on altaz points. You can choose limits in alt and az to eliminate probs with surrounding buildings and also control the density of points.

It will work with TheSky6 and CCDSoft5 and TPoint. You can download an app from Ron Wadowski that will put it all together or you can do it manually. Either way when it is configured it should be able to waltz around the sky mapping while you cool your heels.

jase
10-05-2007, 01:34 PM
:) Good to hear you're back up and running Monte. Wasn't the link I provided in my previous response was it? :lol:

montewilson
10-05-2007, 08:50 PM
Hi Jase - Yes your link popped up a bit later in the dialogue. Below is the reply from Software Bisque. Yours and theirs came up on the same day so I didn't get back to you for a while. Anyway you were spot on. Thanx!

Regards

Monte




Monte,
Thanks for the suggestions.
Are you familiar with Automapper II?
See:
http://www.bisque.com/tom/tpoint/automap.asp (http://www.bisque.com/tom/tpoint/automap.asp)
for more information.
Daniel R. Bisque
Software Bisque
912 Twelfth St.
Golden, CO 80401
Web: http://www.bisque.com

Vince G
11-05-2007, 04:24 PM
Roger, it shouldn't make any difference if the target is out of your field of view when doing a plate solve.
If the solve fails, it may be caused by the star catalogue your using, or it's outside the pointing error setting in CCDCommander.
I've had up to a 1200 arc sec pointing error with my LX200GPS, but it still plate solved.
If you have any questions I'll be happy to help.
Regards
Vince G